"DingTalk has created 'silicon-based employees'"
```
Author | Chai Xuchen
Editor | Zhou Zhiyu
Just as Doubao Assistant enters mobile phones, DingTalk follows with the “D Plan,” bringing the future where AI works for people right before everyone’s eyes.
Recently, DingTalk released exclusive hardware for enterprises, Ding Talk Real, along with an AI system called Agent OS. CEO Chen Hang stated, "In the future, all AI Agents on DingTalk will be built and run based on Agent OS, allowing AI to directly connect with the physical world."
DingTalk harbors vast ambitions, hoping to lead its ecosystem partners to bid farewell to the application architecture of the mobile internet era. By developing the operating system for the AI era, it aims to make "silicon-based employees" working possible.
At this year’s OpenAI developer conference, Sam Altman predicted that the first truly meaningful Agents would integrate into enterprise systems, becoming a "new job category" participating in output. Human responsibilities would shrink to decisions and instructions, while execution is left for machines.
However, a mere chat box cannot carry such ambition. To turn AI capabilities into productivity, a unified AI system is required to coordinate fragmented models and platforms, truly unleashing productivity.
As a crucial entry point for Alibaba's AI strategy on the B-end, DingTalk aims to unify AI capabilities at this critical moment, to become the Android and iOS of the AI era, and make AI capability truly “monetizable.”
Igniting AI Productivity
2025 is the year of explosive AI application, and while the world knows the AI wave has arrived, there is chaos on the ground.
On the one hand, AI is more powerful than ever—capable of writing reports, making videos, coding, and forecasting. Yet the blooms of various AI applications leave enterprises scrambling between platforms. Furthermore, AI tools are isolated; data does not communicate, intentions do not coordinate, and individual Agents can only run on specific platforms.
This “island state” is much like the early days of PC software, incompatible with each other.
Yet a truly intelligent productivity tool should allow users to say, “Help me prepare next week’s report to the boss,” and have AI automatically retrieve project data, create the PPT, book the meeting room, notify participants—all without user intervention.
This requires a unified scheduling hub, not just more standalone tools. Clearly, the AI era also needs an ecosystem platform to coordinate Agents for collaborative production.
As the B-end steward of productivity, DingTalk has made the first move, firing the starting shot.
On December 23, DingTalk released more than 20 AI-native products, the star of which is a new intelligent infrastructure built with “Agent OS” as the core and “Ding Talk Real” as the terminal. This enables large models, AI tools, and processes to operate, coordinate, and dispatch within Agent OS like Apps—liberating humans from repetitive workflows.
With Agent OS collaboration, DingTalk has opened up numerous frequently-used, reusable enterprise Agent scenarios, covering key business areas like AI travel, AI recruitment, and AI customer service. No longer simply iterating as an office software, DingTalk is now building itself as a full system designed for AI.
And all these features are to serve real productivity needs.
Previously, Chen Hang noticed that the industry, including the DingTalk team, still regarded AI as a tool—writing weekly reports, making PPTs, generating images and text—but this does not touch the core. He believes that the AI era must change the way work is done.
“Previously, humans were the main executors and thinkers, with AI as an assistant; the new DingTalk aims to make AI the main body, with nearly infinite memory and global computational abilities, able to perceive, judge, plan, and execute proactively, while humans step back to be decision-makers and commanders, making key judgments based on AI results,” says Chen Hang.
It can be said that DingTalk wants to upgrade AI from “scattered capabilities” to “systematic productivity.” If Windows enabled everyone to use a computer, Agent OS will let every enterprise use systematic AI.
But simply connecting Agents is not enough—AI also needs data precipitation, and hardware is its entry point to obtain data and connect with the real world.
Ding Talk Real arrived accordingly. DingTalk describes it as the key extension of Agent OS in the physical world, just like AI’s body. “The phones and computers we use daily are designed for humans; Ding Talk Real is a device designed specifically for Agents,” says Chen Hang.
In DingTalk's vision, it can let DingTalk keep working for you on your computer, accessing all your data in the internal environment, solving problems and finding opportunities 24/7 without interruption. Enterprises only need to deploy Ding Talk Real locally, and employees can remotely call Agents through DingTalk from anywhere.
Through this “software-hardware integration” strategy, DingTalk intends to build a new moat in the B-end market. Pure software SaaS services are easily replaceable, but once DingTalk Real is deployed, and business processes, physical devices, and Agent OS are deeply bound, switching costs rise exponentially. This physical barrier surpasses even “user habits.”
Alibaba’s B-end Lightning War
At this moment, DingTalk has officially left behind its mobile internet application form. Chen Hang says, “Today, we can announce DingTalk has fundamentally changed, transforming into an AI operating system.”
Many ask, as a super App with 700 million users, why does DingTalk risk so much in making hardware, developing an OS, and restructuring so fundamentally? The answer is simple: in the AI era, old maps won’t find new continents. In the battle of entry points among internet giants, DingTalk holds the fort for Alibaba on the B-end track.
Public data shows that by 2025, China’s collaborative office market will exceed 30 billion yuan. DingTalk, WeCom (WeChat Work), and Feishu hold over 75% of the share, with DingTalk ranked first at about 32.7%; WeCom second at about 23.4%; Feishu third at about 18.9%.
It is clear that DingTalk's once dominant position has been eroded. Even Alibaba-invested AI startups like MiniMax, Moonshot AI, and Zhipu AI have turned to Feishu. “Feishu’s multi-dimensional tables are at least 12 months ahead of DingTalk,” said Feishu CEO Xie Xin in an interview this July.
DingTalk has reached a breaking point, but with Alibaba's overall AI strategy, the tide is turning.
Over the past two years, Alibaba has championed the slogan “Let AI penetrate all industries,” wanting AI to redo every business, redefine efficiency, much like when the internet arrived.
Alibaba CEO Wu Yongming said internally that mobile internet’s biggest value was redistributing people’s attention economy, but it has a limit, with only 24 hours in a day per person; with AI, the biggest difference is a massive boost in human productivity.
The AI wind is blowing. With the shift from Q&A to hands-on work, DingTalk may seize new opportunities in a 700 million user office environment, becoming Alibaba’s new “ticket” into the AI era. With access to the capillaries of the real economy, DingTalk can find more growth and commercialization opportunities, being the ever-present “screw” in every segment of the market supply side.
Hence, Chen Hang returned to DingTalk, choosing to “redo everything” to regain lost ground.
But it's not easy. In the consumer (C-end) market, AI can spread through chat interfaces, but the B-end market is more ruthless. Enterprises do not pay for “intelligence,” only for “certain results”—boosting productivity is the top priority.
Thus, DingTalk's greatest value is fitting the spirit of the times, enhancing user and enterprise productivity, and turning users into “super individuals.” The new DingTalk uses AI to redistribute time and attention, letting people focus on key decisions and high-value output, reclaiming space in the pace of work.
“With stronger data and environments, eventually software itself will disappear—people only need the best deliverables,” asserts Chen Hang. “Whoever first helps AI agents understand industry and enterprise will surely rise in the productivity revolution.”
The new strategy is already showing results. By December, DingTalk’s daily AI calls increased fivefold in a year; the number of Agents created by enterprises and developers on the platform tripled in a year.
“Before August, only 10% believed in DingTalk’s AI transformation. Now, 30% do,” says Chen Hang, describing it as a snowball effect: first, the believers start running, using DingTalk’s new AI products and approaches to spearhead progress.
This is a battle that DingTalk must not lose. If successful, it will redefine industry delivery standards, turning DingTalk from an App into an irreplaceable “infrastructure” in enterprise offices, thus firmly securing Alibaba’s dominance in the B-end market.
Rebuilding Alibaba with AI is the anchor for reassessing Alibaba’s value, and DingTalk is the crucial link for commercialization and real-world scenarios. This elimination race is just beginning; DingTalk needs to accelerate.
Risk Warning and DisclaimerThe market has risks, investment needs caution. This article does not constitute personal investment advice and does not consider the special investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article fit their specific situation. If you invest accordingly, you are responsible for your own risk. ```